Multimodal foundation models are rapidly emerging as a transformative force in enterprise imaging—but their underlying mechanics and clinical implications often remain misunderstood. This session delivers a technical deep dive into how these models integrate visual, textual, and structured data to advance diagnostic imaging. Attendees will explore cutting-edge research, learn about the architectural foundations of state-of-the-art (SOTA) models, and examine real-world use cases in medical imaging AI. Whether you’re a researcher, clinician, or imaging IT professional, this session will help you connect technical innovation with clinical relevance and anticipate where this technology is headed next. 

Additionally, the session will highlight emerging trends beyond multimodal foundation models, specifically addressing the transformative potential of agentic AI and AI agents. Participants will discover how these advanced, goal-driven AI entities differ from traditional AI workflows and examine their potential applications in medical imaging.

Objectives 

• Describe the architectural components and core principles of multimodal foundation models.
• Discuss current state-of-the-art research approaches and their application to medical imaging.
• Explain the clinical and operational potential of multimodal AI in enterprise imaging.
• Describe how agentic AI and AI agents differ from traditional AI workflows, identifying potential use cases and implications for the future of medical imaging informatics.